a3-highgpu-4g as a GKE node
104 vCPU936 GiBx86_644 GPUGoogle Cloud us-central1A3 Highgpu 4g
Where the a3-highgpu-4g goes
Default GKE setup: GKE ≤ 1.36.
CPU103.47 of 104 vCPU left for your pods
Memory910.13 GiB of 936 GiB left for your pods
your podsDaemonSets (200m, 300 MiB assumed)kubelet reservedeviction
110max pods per node
$3,692.92on demand / month · $5.06/h
$2,100.70spot / month (varies)
a3-highgpu-4g GKE max pods and allocatable memory — quick answers
- How many pods can a a3-highgpu-4g run as a node with GKE ≤ 1.36? Up to 110; kubelet keeps 26192 MiB of memory and 330m CPU, leaving 910.42 GiB and 103.67 vCPU for pods.
- How many pods can a a3-highgpu-4g run as a node with GKE 1.37+ (Container-Optimized OS)? Up to 110; kubelet keeps 2250 MiB of memory and 330m CPU, leaving 933.80 GiB and 103.67 vCPU for pods.
- How many 500m / 1 GiB pods fit? 107 per node after DaemonSets; 1 nodes for 10 replicas, about $3,692.92/month on demand.
What kubelet reserves on a3-highgpu-4g
| Setup | Max pods | kube-reserved CPU | Memory reserved + eviction | Allocatable CPU | Allocatable memory | Memory usable |
|---|---|---|---|---|---|---|
| GKE ≤ 1.36 tiered memory reservation | 110 | 330m | 26092 Mi + 100 Mi | 103.67 | 910.42 GiB | 97.3% |
| GKE 1.37+ (Container-Optimized OS) 15 MiB × max pods + 500 MiB, CPU capped at 1 vCPU | 110 | 330m | 2150 Mi + 100 Mi | 103.67 | 933.80 GiB | 99.8% |
How these numbers are computed: reservation formulas. Check a live node with kubectl describe node → Allocatable.
How many pods fit on one a3-highgpu-4g
Assumes DaemonSets (CNI, kube-proxy, log and metrics agents) request 200m CPU, 300 Mi and 3 pods per node. Try your own numbers in the node calculator.
| Pod requests (CPU / memory) | GKE ≤ 1.36 | GKE 1.37+ (Container-Optimized OS) | $ per pod / month |
|---|---|---|---|
| 100m / 128 Mi | 107 | 107 | $34.51 |
| 250m / 256 Mi | 107 | 107 | $34.51 |
| 250m / 512 Mi | 107 | 107 | $34.51 |
| 500m / 512 Mi | 107 | 107 | $34.51 |
| 500m / 1024 Mi | 107 | 107 | $34.51 |
| 1000m / 1024 Mi | 103 | 103 | $35.85 |
| 1000m / 2048 Mi | 103 | 103 | $35.85 |
| 2000m / 4096 Mi | 51 | 51 | $72.41 |
| 4000m / 8192 Mi | 25 | 25 | $147.72 |
Similar GKE node types
| Type | vCPU | GiB | Alloc CPU | Alloc memory | Max pods | $/month | $ per alloc GiB |
|---|---|---|---|---|---|---|---|
| a2-highgpu-8g | 96 | 680 | 95.69 | 659.54 GiB | 110 | $4,318.55 | $6.55 |
| a2-megagpu-16g | 96 | 1360 | 95.69 | 1,325.94 GiB | 110 | $6,421.79 | $4.84 |
| a2-ultragpu-8g | 96 | 1360 | 95.69 | 1,325.94 GiB | 110 | $6,661.79 | $5.02 |
| g2-standard-96 | 96 | 384 | 95.69 | 369.46 GiB | 110 | $2,571.80 | $6.96 |
| a2-ultragpu-4g | 48 | 680 | 47.81 | 659.54 GiB | 110 | $3,330.90 | $5.05 |
| a3-highgpu-2g | 52 | 468 | 51.80 | 451.78 GiB | 110 | $1,846.46 | $4.09 |
| a3-edgegpu-8g-nolssd | 208 | 1872 | 207.41 | 1,827.70 GiB | 110 | $6,905.84 | $3.78 |
| a3-edgegpu-8g | 208 | 1872 | 207.41 | 1,827.70 GiB | 110 | $7,385.84 | $4.04 |
Closest match on other clouds
| Type | vCPU | GiB | Alloc CPU | Alloc memory | Max pods | $/month | $ per alloc GiB |
|---|---|---|---|---|---|---|---|
| g7e.24xlarge | 96 | 1024 | 95.69 | 1,016.86 GiB | 632 | $12,097.68 | $11.90 |
| g6e.24xlarge | 96 | 768 | 95.69 | 762.37 GiB | 492 | $10,997.88 | $14.43 |
| p3dn.24xlarge | 96 | 768 | 95.69 | 759.74 GiB | 737 | $22,784.76 | $29.99 |
| p4d.24xlarge | 96 | 1152 | 95.69 | 1,143.74 GiB | 737 | $16,029.08 | $14.01 |
| p4de.24xlarge | 96 | 1152 | 95.69 | 1,143.74 GiB | 737 | $20,036.35 | $17.52 |
| p2.16xlarge | 64 | 732 | 63.77 | 728.97 GiB | 250 | $10,512.00 | $14.42 |